Hmm, with object-shaders I mean shaders which are run for each pixel drawn of an object, like a common normalmapping shader. For this case, it could be that the same pixel gets attribution from more than one object, if the objects cover that very pixel, but have different depths. With post-processing shaders I meant shaders that work on a G-Buffer or general screen-space bitmaps (like the already rendered image) and are not driven by the actual scene geometry.